Separating drawings, formula and text from free handwriting

This paper describes a method for separating on-line handwritten patterns into characters, figures, and formulas. Today, Tablet PCs and electronic whiteboards provide much larger writing area for pen interfaces unlike PDAs, so that users can easily input text, write mathematical formulas and draw figures on the screen. The fact that people can write these objects by a single pen (marker) without switching the device, mode, software or whatever else and without any writing restriction such as grids or boxes is one of the most important benefits of the pen interfaces but it requires difficult task of segmenting these objects. We apply a probabilistic model for this problem and employ stroke features, stroke crossings and stroke densities. Moreover, we partially apply the approach of segmentation by recognition. Although the current recognizer for formulas is not a true recognizer, we have achieved about 81% correct segmentation for all the strokes in our newly prepared database of the mixed patterns from these objects.